Identification of the Interaction Domain of the Small Terminase Subunit

Institute of Clinical and Molecular Virology. § Hannover Medical School. 1 Abbreviations: HCMV, human cytomegalovirus; BAC, bacterial artificial chro...
5 downloads 0 Views 502KB Size
Download PDF
Biochemistry 2006, 45, 8855-8863

8855

Identification of the Interaction Domain of the Small Terminase Subunit pUL89 with the Large Subunit pUL56 of Human Cytomegalovirus† Corina Thoma,‡ Eva Borst,§ Martin Messerle,§ Manuela Rieger,‡ Jae-Seon Hwang,‡ and Elke Bogner*,‡ Institute of Clinical and Molecular Virology, Schlossgarten 4, 91054 Erlangen, Germany, and Institute of Virology, HannoVer Medical School, Carl-Neuberg-Strasse 1, 30625 HannoVer, Germany ReceiVed January 13, 2006; ReVised Manuscript ReceiVed May 16, 2006

ABSTRACT: The small terminase subunit pUL89 of human cytomegalovirus (HCMV) is thought to be required for cleavage of viral DNA into unit-length genomes in the cleavage/packaging process. Immunoprecipitations with a UL89-specific antibody demonstrated that pUL89 occurs predominantly as a monomer of approximate Mr 75.000 together with a dimer of approximate 150.000. This was confirmed by gel permeation chromatography. In view of its putative function, pUL89 needs to be transported into the nucleus. By use of laser scanning confocal microscopy, pUL89 was found to be predominantly localized throughout the nucleus and in particular in viral replication centers of infected cells. By immunofluorescence, we demonstrated that both terminase subunits co-localized in viral replication centers. Furthermore, analysis with pUL89 GST-fusion protein mutants showed that amino acids 580-600 may represent the interaction domain with pUL56. To verify this result, a recombinant HCMV genome was constructed in which the UL89 open reading frame was disrupted. By transfection of the deletion BACmid alone, we showed that it has a lethal phenotype. Cotransfection assays demonstrated that, in contrast to pUL89 wild-type, a plasmid construct encoding a pUL89 variant without aa 580-590 as well as one encoding a variant without aa 590-600 could not complement the HCMV-pUL89 null genome, thus, suggesting that the 20 aa sequence GRDKALAVEQFISRFNSGYIK is sufficient for the interaction with pUL56 and in conclusion required for DNA packaging.

Human cytomegalovirus (HCMV)1, one of eight human herpesviruses, can cause serious, life threatening diseases. Maturational events of HCMV DNA replication and capsid assembly require efficient translocation of viral gene products into the nucleus of infected cells. Several gene products of HCMV have been identified that appear to be involved in the process of DNA replication, cleavage, and packaging (1). DNA replication leads to the formation of head-to-tail linked concatemeric DNA which must be resolved into unit-length genomes during packaging. Proteins involved in viral DNA cleavage and packaging are so-called terminases. Terminases were first described as specific enzymatic proteins for dsDNA bacteriophages with multiple activities required for the whole packaging process (2-4). Terminases are heteromultimers composed of a small subunit that binds the DNA † This work was supported by the DFG (Deutsche Forschungsgemeinschaft; BO1214/5-4). E.B. is a recipient of a HWP fellowship from the Friedrich-Alexander University of Erlangen-Nu¨rnberg. * Corresponding author. Mailing address: Institute of Clinical and Molecular Virology, Schlossgarten 4, 91054 Erlangen, Germany. Phone: +49-9131-8522104. Fax: +49-9131-8526493. E-mail: [email protected]. ‡ Institute of Clinical and Molecular Virology. § Hannover Medical School. 1 Abbreviations: HCMV, human cytomegalovirus; BAC, bacterial artificial chromosome; GST, glutathione-S-transferase; FCS, fetal calf serum; DMEM, Dulbecco’s modified Eagle’s medium; GFP, green fluorescent protein; IPTG, isopropyl β-D-1-thiogalactopyranoside; MOI, multiplicity of infection; PAGE, polyacrylamide gel electrophoresis; PAA, phosphonoacetic acid; PMSF, phenylmethanesulfonyl fluoride; ORF, open reading frame.

and a large subunit that has, besides an endonuclease activity, an ATPase activity mediating the translocation of DNA into the preformed capsid (5-13). Our previous results together with the observations from Underwood et al. (14) and Krosky et al. (15) indicate that the terminase of human cytomegalovirus consists of the highly conserved proteins pUL56 and pUL89 (16-21). While it is supposed that the large terminase subunit pUL56 (i) mediates the specific binding to packaging elements, socalled pac motifs, on the concatamers; (ii) provides energy for translocation of the DNA to the procapsids; and (iii) associates itself with the capsid for enabling the entry of the DNA, the small subunit pUL89 seems to be mainly required for cleavage of concatemeric DNA into unit-length genomes (20). Both proteins are highly conserved throughout the herpesviruses. Recently, we demonstrated by structural analysis of single particles that pUL56 forms dimers of two open rings, a characteristic of DNA metabolizing proteins (22). Adelman et al. (23) demonstrated that the HSV-1 homologue of pUL56, pUL28, plays a role in DNA packaging as a sequence-specific DNA-binding protein. Despite the importance of these proteins and the intensive research into biochemical and functional aspects of terminases, there is still a lack of information regarding their intracellular localization. In this study, we characterized the small terminase subunit pUL89 of HCMV. We could demonstrate that (i) the protein is expressed as a 75 kDa protein and has a tendency to dimerize, (ii) it is translocated in infected cells

10.1021/bi0600796 CCC: $33.50 © 2006 American Chemical Society Published on Web 06/22/2006

8856 Biochemistry, Vol. 45, No. 29, 2006

Thoma et al.

Table 1: Oligonucleotide Primers Used in PCR

into viral replication centers, (iii) aa 580-600 are required for the interaction with the large terminase subunit pUL56, and (iv) a UL89 construct without this region was unable to complement a UL89-null genome. MATERIAL AND METHODS Cells and Viruses. Human foreskin fibroblasts, HFF-cells, at passage 10-20 were grown in DMEM supplemented with 10% fetal calf serum (FCS), 2 mM glutamine, penicillin (5 U/mL), and streptomycin (50 µg/mL). Preparation of HCMV AD169 and infection of HFF at an MOI of 3 was carried out as described before (24). Inhibition of the viral polymerase was achieved by addition of PAA (200 µg/mL) during the entire infection. Insect cells 5B1-4 (High five) were grown in TC-100 medium supplemented with glutamine, gentamicin (60 µg/ mL), and 10% fetal calf serum. Recombinant baculovirus expressing rpUL89 containing an N-terminal Anti-Xpress antibody epitope was generated by site-specific transposition of the expression cassette into the bacmid bMON14272 as described by Life Technology. BAC Mutagenesis. For generation of HCMV-GFP-BAC∆UL89, exon 2 of ORF UL89 (nt 133449-134499) was deleted in the parental BAC HCMV-GFP by the ET mutagenesis procedure using linear PCR fragments (25). A PCR fragment was generated using the following oligonucleotides: UL89.for, 5’-ACCCTGAACGGATGGCGTGTATATCGTCACACAGGTAGGTGGCCATGATGCCCCGAAAAGTGCCACCTGCAGAT-3’; UL89.rev, 5’-TCTGTCTCCTCTCTACACAGAGCATCCGAGGACAAAACTTCCACTTGCTGGTGACACAGGAACACTTAACGGCTGA3’containing approximately 20 nt homologous to the kanamycin resistance gene of pori6K-F5 (26) and 50 nt homologous to the regions upstream and downstream of UL89 exon 2. The resulting fragment was introduced into Escherichia coli containing the parental HCMV-GFP-BAC for homologous recombination as previously described (27). Bacterial clones were selected on agar plates containing 25 µg/mL kanamycin plus 17 µg/mL chloramphenicol and checked for the presence of mutant BACs by restriction analysis and agarose gel electrophoresis. The kanamycin resistance marker was subsequently excised by site-specific recombination in E. coli using Flp recombinase (28-29). BAC Isolation. BAC DNA was obtained from 500-mL overnight cultures using Nucleobond PC500 columns (Macherey Nagel, Du¨ren, Germany) according to the instructions of the manufacturer.

Immunofluorescence and Immunoprecipitation. For immunofluorescence, HFF cells were grown on cover slips. At the appropriate time point, cells were fixed with 3% paraformaldehyde as described previously (30). Detection of HCMV pUL89 was carried out with pAbUL89 (19) for 45 min at room temperature prior to further incubation for 45 min with Cy3-labeled goat anti-human F(ab’)2 fragments. For double staining, HCMV pUL89specific human pAbUL89 and IE1-specific monoclonal antibody mAb pp63-27 (provided by T. Stamminger, 31) was used in the first incubation. In the second incubation, Cy3-labeled anti-human and Cy2-labeled anti-mouse F(ab’)2 fragments were used. For immunoprecipitation, infected or mock-infected cells (T-75 cm2 flasks; 2 × 106 cells) were radiolabeled 60 h postinfection with 50 µCi/mL [35S]methionine for 12 h. Total cell extracts were prepared from labeled cultures by solubilization in immunoprecipitation buffer (20 mM Tris-HCl, pH 7.5, 100 mM NaCl, 1% NP-40, 5 mM EDTA, 25 mM iodacetamide, 0.4% sodiumdeoxycholate, 1 mM PMSF, and 100 U/mL Trasylol) and ultrasonic treatment. Insoluble material was sedimented for 30 min at 100 000g. Comparable amounts of extracts and pUL89-specific antibody (1:20) were used for precipitation as described previously (16). Plasmid Construction and Oligonucleotides. Plasmid GSTUL89 was generated by digestion of pcDNA-UL89 (20) and pGEX-5X-1 (Amersham Pharmacia Biotech, Freiburg, Germany) as described previously (19). The 2.025 bp fragment encoding the HCMV pUL89 was ligated in-frame into pGEX-5X-1, yielding pGEX-UL89. Oligonucleotide-directed mutagenesis of GST-UL89 was performed. Constructs GST-UL89-1, GST-UL89-2, GSTUL89-3, GST-UL89-4, and GST-UL89-5 were generated using the QuickChangeTM site-directed mutagenesis kit (Stratagene) as recommended by the supplier and pairs of synthetic oligonucleotides as described in Table 1 (construct GST-UL89 served as template). The construct GST-UL89-5 with overlapping regions between GST-UL89-3 and GSTUL89-4 was chosen based on computer analysis that suggests aa residues 580-600 are exposed at the surface of the protein. All mutations were verified by DNA sequencing. For construction of plasmids pUC-UL89P, pUC-UL89P∆10aa1, pUC-UL89P∆10aa2, and pUC-UL89P∆9aa, respective fragments were obtained using cosmid pCM1106 as a template for PCR and ligation into pUC-18. The pairs of synthetic oligonucleotides are described in Table 2.

Interaction Domain HCMV pUL89

Biochemistry, Vol. 45, No. 29, 2006 8857

Table 2: Oligonucleotide Primers Used in PCR

Table 3: pUL89 Mutants Used in this Study

construct

DNA sequence w/o GST

amino acid sequence w/o GST

Predicted Mr with GST

GST-UL89 GST-UL89-1 GST-UL89-2 GST-UL89-3 GST-UL89-4 GST-UL89-5

1-2025 bp 1-600 bp 541-1200 bp 1171-1800 bp 601-2025 bp 1741-1800 bp

1-675 1-200 180-400 390-600 200-675 580-600

101.000 48.000 50.200 49.100 78.250 28.200

Protein Purification. A fresh overnight culture of E. coli BL21 carrying the plasmid pGEX-UL89, pGEX-UL89-1, pGEX-UL89-2, pGEX-UL89-3, pGEX-UL89-4, pGEXUL89-5, or pGEX-5X-1 was diluted 1:100 in LB medium (1 L) containing 50 µg/mL ampicillin and incubated at 37 °C. After the cells reached an OD600 of 0.5, the GST-protein expression was induced by addition of IPTG to a final concentration of 0.1 mM and incubated for 2 h at 37 °C. Sedimented cells were lysed in 40 mL binding buffer (20 mM sodium phosphate, 0.15 M NaCl, pH 7.4) containing 250 µL 1 M MgCl2, 25 µL 1 M MnCl, 40 µL of DNase I (10 mg/mL), and 400 µL of lysozyme (10 mg/mL), incubated on ice for 30 min, and sonicated on ice. Undissolved material was sedimented for 40 min at 5 000g at 4 °C and passed through a 0.45 µm filter. The purification was performed with a 1 mL GSTrap column by using an A ¨ kta Prime machine (Amersham Pharmacia Biotech) according to the instruction of the manufacturer. The column was washed with 3 bed volumes of binding buffer prior to loading the proteins. Elution was performed with 10 bed volumes of elution buffer (50 mM Tris-HCl, 10 mM glutathione, pH 8.0), and 10 fractions were collected. The fractions were analyzed by SDS-PAGE. Fractions containing the GST-UL89 proteins (Table 3) were stored at -80 °C. Protein Purification of Recombinant, BaculoVirus-Expressed rpUL89. High-five cells (4 × 108) expressing the recombinant pUL89 (20) were harvested 48 h post-infecton. Sedimented cells were lysed in 50 mL of cation exchange buffer (20 mM MES, pH 6.5, 150 mM NaCl, and protease inhibitors) and sonicated on ice. Cell lysates were sedimented and passed through a 0.2 µm filter prior to loading onto an equilibrated anion exchange column (HiTrap Q HP, 1 mL bed volume; Amersham Bioscience). The purification was performed at 4 °C by using an A ¨ KTA Prime (Amersham

Bioscience) as described before (20). Fractions containing the protein were subjected to a second chromatographic step using a molecular size exclusion column (HiLoad 16/60 Superdex 200 prep grade) and an A ¨ kta FPLC (Amersham Bioscience) as described (20). Molecular mass markers, thyroglobulin (669 kDa), ferritin (440 kDa), catalase (232 kDa), and aldolase (158 kDa) were analyzed by gel permeation chromatography as well. In Vitro Translation. Plasmid pcDNA-UL56 (0.5 µg) and plasmid pHM123 (0.5 µg), encoding HCMV IE1, were incubated with [35S]methionine (10 mCi/mL) and 20 µL of TNTT7 Quick Master Mix (Promega, Madison, WI) in a final volume of 30 µL for 1 h at 30 °C. Translation products were analyzed by SDS-PAGE. SDS-PAGE and Autoradiography. The proteins of the supernatants or precipitated proteins were separated on 10% polyacrylamide gels, transferred to nitrocellulose sheets, and subjected to Western blot analysis or stained with coomassie brilliant blue. The antibody pAbUL89 (1:20) specific for pUL89 was used as the primary antibody prior to incubation with horseradish peroxidase-conjugated anti-human F(ab’)2 fragments (1:500). After immunostaining, radioactive blots or gels were subjected to autoradiography. In Vitro Binding Assay. For in vitro binding analysis, equal amounts of GST fusion proteins (see above) loaded on glutathione-sepharose 4B (Amersham Biosciene) were incubated with in vitro translated pUL89 or IE1 for 2 h at 4 °C in 500 µL of binding buffer (0.05% Nonidet P-40, 50 mM Hepes, 10% glycerol, 0.1% BSA, and 300 mM NaCl, pH 7,3). Samples were washed with binding buffer and subsequently subjected to SDS-PAGE, consecutive fixation, and autoradiography. For quantification of radioactive signals, a BioImager (Raytest, Germany) was used. PreScission Protease Treatment. To cleave the GST-tag from GST-UL89, after binding to the GSTrap column, the protein was incubated with 80 U PreScission protease (Amersham Bioscience) for 12 h at 4 °C under rotation. After removal of the GST-tag from GST-UL89 during incubation with the protease, the resulting proteins were eluted in 10 fractions (1 mL) of PreScission buffer (50 mM Tris, pH 7.0, 150 mM NaCl, 1 mM EDTA, and 1 mM DTT). The fractions containing the protein (0.97 mg/mL) were stored at -80 °C. Protein/Protein Interaction by Chemical Cross-Linking. pUL89 (300 ng) was incubated for 5 min at room temperature

8858 Biochemistry, Vol. 45, No. 29, 2006

Thoma et al.

FIGURE 1: Identification of HCMV pUL89 in infected HFF cells. (A) Lysates of mock-infected (lane 1) or HCMV-infected HFF cells (lane 2) at 72 h post-infection were subjected to 8% SDS-PAGE prior to transfer onto a nitrocellulose membrane. Immunoblot analysis was performed using the pUL89-specific antibody pAbUL89. (B) HCMV infected (lane 2) or mock-infected cells (lane 1) were radiolabeled 60 h post-infection with 50 µCi/mL [35S]methionine for 12 h. At 72 h post-infection, cells were immunoprecipitated with pAbUL89 prior to 8% SDS-PAGE and autoradiography. (C) Analysis of the association of pUL89 with itself by protein cross-linking using a bacterial fusion protein. The purified bacterial protein pUL89 was either untreated or reacted with glutaraldehyde for 10 min, followed by SDS-PAGE and Western blot analysis.

in PBS in a final volume of 20 µL. The cross-linking reaction was started by addition of 0.1% (v/v) glutaraldehyde to a final concentration of 0.01%. The reaction was stopped by addition of 10 µL of SDS-loading buffer and denaturation of the samples for 2 min at 95 °C. The proteins were separated by 8% SDS-PAGE and analyzed by Western blot with the specific antibody pAbUL89. Cotransfection Assays. For transient complementation of the HCMV pUL89-null genome, 1 µg of HCMV-GFP-BAC∆UL89, together with 2 µg off either pUL89 expression plasmid pUC-UL89P, pUC-UL89P∆10aa1 (∆ aa580-590), pUC-UL89P∆10aa2 (∆ aa590-600), or pUC-UL89P∆9aa (∆ aa100-109), was transfected into hTERT-RPE1 cells (Clontech) using the Transfectin reagent (Bio-Rad) according to the manufacturer’s instructions. Ten days after transfection, the cells were analyzed for plaque formation using fluorescence microscopy. The parental HCMV-GFP-BAC and HCMV-GFP-BAC-∆UL89 transfected together with HCMV cosmid clone pCM1106 (32) comprising the UL89 ORF plus flanking homologous regions were used as positive controls. RESULTS Characterization of pUL89 in Infected Cells. To determine the expression of pUL89 in infected cells, immunoblots with mock and HCMV-infected cells were performed. The monospecific antibody pAbUL89 (19), which was purified from hightiter human serum by column affinity chromatography (Affigel 10/15-pUL89), recognized a 75 kDa protein in HCMV-infected cells (Figure 1A, lane 2), that corresponds with the predicted size of the amino acid sequence. In the control, no protein was detected (Figure 1A, lane 1). To further investigate the presence of pUL89 in infected cells, immunoprecipitation with pAbUL89 was performed prior to SDS-PAGE and autoradiography. Interestingly, the 75 kDa protein and an additional band of approximately 150 kDa were detected in infected cells (Figure 1B, lane 2). It is

FIGURE 2: Purification of pUL89. (A) Gel permeation chromatography of rpUL89. The protein was subjected to chromatography through a HiLoad 16/60 Superdex 200 prep column using an A ¨ KTA Explorer. The flow-through was recorded at OD280. The column was calibrated with the molecular weight standards aldolase (158 kDa), catalase (232 kDa), ferritin (440 kDa), and thyroglobulin (669 kDa) as indicated by arrows. (B) Analysis of the peak fractions by urea gel electrophoresis. The fractions f1-f5 (corresponding to 3542.5 mL), f16/17/18 (corresponding to 58-62.5 mL), and f24/25 (corresponding to 70-73 mL) were separated on the gel prior to immunoblot analysis with monoclonal antibody against the express tag of rpUL89. (C) Analysis of fractions 16, 17, and 18 by 8% SDS-PAGE and Western blot analysis using the anti-Xpress antibody. The molecular weight markers (M) are indicated on the left.

reasonable to speculate that the latter protein may represent a dimer of pUL89 and that pUL89 also forms a dimer as the large terminase subunit pUL56. Therefore, the interaction of pUL89 with itself was examined by chemical protein cross-linking using a bacterial fusion protein. The GST-UL89 fusion protein was purified including cleavage of the GSTtag and subjected to treatment with glutaraldehyde for 0 and 10 min before the proteins were separated by SDS-PAGE and analyzed by immunoblotting. Again, a new protein band appeared at 150 kDa, indicative of the dimer (Figure 1C, lane 2). To further analyze the molecular size of pUL89, the protein was expressed as a recombinant protein utilizing a baculovirus system and subjected to molecular sizeexclusion chromatography together with marker proteins (Figure 2A). The peak fractions f1-f5 (corresponding to 3542.5 mL), f16/17/18 (corresponding to 58-62.5 mL), and f24/25 (corresponding to 70-73 mL) were analyzed by urea gel electrophoresis followed by Western blot (Figure 2B). Interestingly, the only observed rpUL89 was in fraction 1618, thus, indicating that pUL89 neither forms high molecular weight complexes (fraction 1-5) nor monomers under native conditions (fraction 24/25). Analysis of fractions 16 and 17 by 8% SDS-PAGE and Western blot analysis demonstrated

Interaction Domain HCMV pUL89

FIGURE 3: Confocal microscopy analysis of the intranuclear distribution of HCMV pUL89. HFF cells were mock-infected (i and j) or infected with HCMV for 12 (a and b), 24 (c and d), 48 (e and f), and 72 h (g and h) and subjected to immunofluorescence with antibodies directed against pUL89 (pAbUL89; panels a, c, e, g, and i) and against the IE1 protein (mAb63-27; panels b, d, f, h, and j). HFF cells infected with HCMV in the presence of PAA were stained with antibodies against pUL89 (k) and against IE1 (l) at 72 h post-infection, respectively.

that the detected protein migrated with an Mr of 150.000, the molecular mass of a dimer (Figure 2C). These results demonstrated that pUL89 is expressed in infected cells as a monomer of 75 kDa with a tendency to dimerize. Intracellular Distribution of pUL89. To examine the precise intracellular distribution of pUL89, immunofluorescence was performed. Coverslip cultures were mock-infected or infected with HCMV (MOI of 1) for 12, 24, 48, and 72 h prior to immunofluorescence and analysis by confocal microscopy. In infected cells, pAbUL89 detected antigen predominantly in the nucleus (Figure 3). Distribution and

Biochemistry, Vol. 45, No. 29, 2006 8859 intensity of fluorescence changed during the course of infection. At 12-24 h post-infection, nuclear staining in prereplication centers was observed, that increased over time until late time points after infection when in addition to homogeneous nuclear staining bright intranuclear patches were detected (Figure 3a,c,e,g). As a control for nuclear distribution, double staining with the antibody mAb pp6327 against IE1 was performed. The IE1 migrates into the nucleus after translation, where it serves as a regulatory protein of the synthesis of early and late HCMV mRNAs. In infected cells, a homogeneous staining of IE1 throughout the nucleus was observed (Figure 3b,d,f,h). To determine whether the translocation of pUL89 is an early event, immunofluorescence was performed in the presence of phosphono acetic acid (PAA; 200 µg/mL), an inhibitor of viral DNA replication. At 72 h post-infection, double staining with antibodies directed against the HCMV pAbUL89 and IE1 proteins, respectively, was undertaken. The immediate early proteins were localized in the nucleus, as expected, as well as pUL89 (Figure 3k). The pattern of the pUL89 immunofluorescence equals that of untreated cells at early times of infection (Figure 3a). These experiments showed that pUL89 is translocated into the nucleus early after infection and that localization in intranuclear patches is dependent on viral DNA replication. Co-Localization of pUL89 with the Terminase Subunit pUL56. To analyze the interaction of pUL89 with the terminase subunit pUL56, laser scanning confocal microscopy was performed. At early times after infection, pUL56 and pUL89 were observed in nuclear spots, representing prereplication centers (Figure 4a-f). At 48 h after infection, both proteins were found to localize to replication centers (Figure 4g-i). The fluorescence pattern changed after 72 h post-infection to brighter staining of the replication centers (Figure 4j-l). Mock-infected HFF revealed no specific immunofluorescence signals (Figure 4m,n). We concluded from these results that both proteins are localized in the viral replication centers as required for DNA-packaging proteins. Identification of pUL89 Domains Required for Interaction with pUL56. To verify a physical interaction of the HCMV terminase subunits, in vitro binding assays were carried out using GST-fusions of the wild-type or of mutant pUL89 proteins. To this end, GST-UL89, GST-UL89-1, GST-UL892, GST-UL89-3, GST-UL89-4, GST-UL89-5 (Figure 5A), or GST itself was immobilized on glutathione-sepharose beads and incubated either with in vitro translated [35S]methionine-labeled pUL56 (Figure 5B, upper panel) or [35S]methionine-labeled IE1 (Figure 5B, lower panel). The amount of bound material was analyzed by SDS-PAGE and autoradiography. As shown in Figure 5B, pUL56 interacted specifically with GST-UL89-3, GST-UL89-4, and GSTUL89-5 but not with GST alone, GST-UL89-1, and only marginally with GST-UL89-2. In vitro translated IE1, used as a negative control, did not interact with the GST fusion proteins. To estimate the intensity of the pUL56-pUL89 interaction, quantification was performed using bioimaging analysis of fluorographs from three different experiments (Figure 5C). Interestingly, the binding of GST-UL89-5 was stronger than GST-UL89-3, but slightly weaker than GSTUL89-4. These experiments led to the suggestion that at least one region necessary for the interaction of pUL89 with pUL56 is located between aa 580 and 600 of pUL89.

8860 Biochemistry, Vol. 45, No. 29, 2006

Thoma et al.

FIGURE 4: Co-localization of pUL89 with HCMV pUL56 by confocal microscopy. Mock-infected (m and n) or infected cells were double-stained with pAbUL89 and pAbUL56 directed against pUL56 at 12 (a-c), 24 (d-f), 48 (g-i), and 72 h (j-l) postinfection. The images represent the central section of the cells analyzed by laser scanning confocal microscopy. A merge of the Cy2- and Cy3-stained cells is shown to the right.

Construction of the Replication-Deficient HCMV-GFPBAC-∆pUL89. To verify the interaction domain of pUL89 with pUL56, an HCMV-GFP pUL89-null mutant was constructed. Since the UL89 gene consists of two exons and overlaps with other ORFs, exon 2 of UL89, which is clearly separated from other ORFs, was deleted in order to minimize any potential effect of the mutation on the neighboring genes (Figure 6A-C). For generation of the mutant BAC, the UL89 exon 2 was replaced by a kanamycin resistance marker. In the resulting HCMV-GFP-BAC-∆UL89, the kanamycin resistance gene was removed by Flp recombinase-mediated recombination. BACs HCMV-GFP-BAC and HCMV-GFPBAC-∆UL89 were digested with HindIII and separated on an agarose gel (Figure 6D). As can be seen in the figure, a band of 7.5 kb characteristic of the parental HCMV-GFPBAC is shifted to a lower molecular weight in HCMV-GFPBAC-∆UL89, giving rise to a double band at 6.5 kb. Analysis of UL89 Deletion Constructs by Complementation Experiments with the HCMV UL89 Null Genome. Cotransfection analyses were performed to analyze the UL89 deletion constructs with the recombinant genome HCMVGFP-BAC-∆UL89. In contrast to HCMV-GFP-BAC, the recombinant bacmid HCMV-GFP-BAC-∆UL89 transfected

FIGURE 5: Identification of the pUL89 domains interacting with pUL56. (A) Schematic drawing of the pUL89 mutant proteins. (B) Glutathione-Sepharose 4B beads loaded with GST, GST-UL89, GST-UL89-1, GST-UL89-2, GST-UL89-3, GST-UL89-4, and GSTUL89-5 were incubated with in vitro translated UL56 or IE1. The bound material was eluted and subjected to SDS-PAGE, prior to autoradiography. The position of pUL56 and IE1 are indicated on the right. (C) Specifically the immobilized radioactivity was quantified using bioimaging analysis. The relative amounts of bound radioactivity are indicated by percent values in the diagram. Error bars on the histogram are ( standard deviation from three independent experiments.

alone led to the formation of single green fluorescent cells only but not to plaque formation, indicating a lethal phenotype (Figure 7A,B). We therefore examined the ability of pUL89 proteins to complement the HCMV UL89-null genome. Cotransfection with HCMV cosmid pCM1106 (32), plasmid pUC-UL89P carrying the authentic pUL89 ORF plus its promoter region, or plasmid pUC-UL89P∆9aa, carrying a deletion at the N-terminus of pUL89, resulted in plaque formation and therefore phenotypic complementation (Figure

Interaction Domain HCMV pUL89

Biochemistry, Vol. 45, No. 29, 2006 8861

FIGURE 6: Construction of a HCMV UL89-null genome. (A) The top line represents the genome structure of the parental HCMVGFP-BAC. (B) The region containing the ORF of UL89 (exons 1 and 2) is expanded. (C) The final BAC HCMV-GFP-BAC-∆UL89 was constructed by replacement of UL89 exon 2 with a kanamycin resistance cassette, followed by its removal by Flp recombinasemediated recombination. (D) DNA of HCMV-GFP-BAC and HCMV-GFP-∆UL89-BAC was digested with HindIII and separated on an agarose gel. The relevant bands characterizing the BACs are indicated by arrows.

7C-E). Cotransfection with constructs containing either a deletion of aa 580-590, pUC-UL89P∆10aa1, or a deletion of aa 590-600, pUC-UL89∆10aa2, were not sufficient for complementation, because only single green cells were observed (Figure 7F,G). These experiments demonstrated that aa 580-600 are required for the function of pUL89 and furthermore for viral replication. DISCUSSION The process of herpesvirus DNA packaging requires the involvement of terminases. These enzymes (i) provide ATP for translocation of the DNA into the preformed capsids, (ii) have the ability to bind to DNA and to capsids, and (iii) have nuclease activity which releases unit-length genomes from concatemeric-replicated DNA. In the case of human cytomegalovirus (HCMV), the terminase is composed of the large subunit pUL56 and the small subunit pUL89 (16, 1920). Both proteins are conserved throughout the herpesviruses and have functions analogous to terminases of doublestranded DNA bacteriophages. In the present work, we characterized the most conserved terminase subunit, pUL89 of HCMV. In infected cells, the UL89 ORF encodes proteins of approximately 75 and 150 kDa. While the prominent 75 kDa protein represents the monomer and is consistent with the predicted molecular

FIGURE 7: Cotransfection analysis with the recombinant genome HCMV-GFP-∆UL89. Transfection with wild-type HCMV-GFPBAC (A) or the recombinant BAC HCMV-GFP-BAC-∆UL89, containing a deletion of exon 2 of UL89 (B), or cotransfection with the cosmid pCM1106 including both exons of pUL89 and flanking homologous regions (C), pUC-UL89P carrying the authentic pUL89 (D), pUC-UL89P∆9aa with a deletion at the N-terminus (E), pUCUL89P∆10aa1 with a deletion of aa 580-590 in UL89 (F), or pUCUL89P∆10aa2 with a deletion of aa 590-600 (G) was performed. Ten days post-transfection, fluorescence microscopy was performed.

weight, the slower-migrating 150 kDa protein is observed

8862 Biochemistry, Vol. 45, No. 29, 2006 after immunoprecipitation with the pUL89-specific antibody. This form of pUL89 was also observed after gel permeation chromatography with marker proteins, thus, indicating that the recombinant rpUL89 forms a dimer. Interestingly, the homologous viral gene in HSV-1, UL15, encodes multiple proteins with apparent molecular weights from 80.000 to 87.000 (33-34). Only the high molecular weight pUL15 protein is a structural protein associated with DNA-containing capsids (C capsids). Some evidence was provided that suggested that the pUL15 forms are bound to the capsids by interactions with the other terminase subunit pUL28 and the portal protein pUL6 of HSV-1 (35). In contrast to these observations, HCMV pUL89 is not a constituent of the virion and therefore a nonstructural protein (data not shown). Our previous data suggested that the association of HCMV pUL89 with the capsid is mediated by interaction with pUL56, but pUL89 was never found together with pUL56 after DNA-cleavage (20). In agreement with the knowledge about the terminases of dsDNA bacteriophages (36), it is reasonable to speculate that, shortly after DNA cleavage, pUL89 is no longer associated with the DNA and is recycled for further cleavage/packaging processes. Furthermore, our findings indicate that pUL89 has a tendency to oligomerize, which is common for terminases of all double-stranded DNA bacteriophages analyzed so far (36). For instance, the small terminase subunit gp16 of bacteriophage T4 forms ring-like structures by oligomerization (37). We investigated the intracellular localization of pUL89 by immunofluorescence analysis of infected cells. At early times after infection (12-24 h post-infection), staining in pre-replication centers was obtained with specific antibody pAbUL89. During infection (from 48 h post-infection onward), the fluorescence pattern changed and pUL89specific signals were increasingly concentrated in replication centers. Experiments in the presence of phosphono acetic acid showed that this localization is dependent on viral replication. Since comparable to the large terminase subunit (20, 38) also the small one is associated with viral replication centers, it is tempting to speculate that DNA packaging is closely associated with replication. Our observations are in line with those from Yu and Weller (34) demonstrating that the HSV-1 homologue pUL15 also localizes to replication centers. Previously, we demonstrated only by co-immunoprecipitation that both terminase subunits interact with each other (19). In the study presented here, the co-localization of pUL56 and pUL89 in infected cells was examined by laser scanning confocal microscopy. We could demonstrate that both proteins were translocated into viral replication centers. These findings are consistent with the reports of nuclear translocation of the homologous HSV-1 proteins (39-40). However, Koslowski et al. (40) as well as Abbots et al. (39) demonstrated that the pUL56 homologue of HSV, pUL28, is only translocated into the nucleus in the presence of pUL15, thus, implicating that in HSV-1, pUL15 is a chaperone for pUL28. These observations are different compared to our data with HCMV. We identified previously a NLS at the carboxy terminus of pUL56, and further evidence was provided that nuclear translocation of pUL56 is mediated by the importin-dependent pathway (41). Furthermore, analyses were performed to identify the interaction domain of pUL89 with pUL56. Since nothing was

Thoma et al. known so far about pUL89 domains, overlapping protein mutants were constructed and analyzed by in vitro binding assays. We were able to identify an interaction domain between aa 580-600 (GRDKALAVEQFISRFNSGYIK). This result was verified by analysis with a recombinant HCMV genome, HCMV-GFP-BAC-∆UL89 carrying a disrupted UL89 gene. We demonstrated by transfection that the BAC HCMV-GFP-BAC-∆UL89 has a lethal phenotype. Cotransfection assays with HCMV cosmid pcm1106, the plasmid containing the full-length UL89 ORF including its promoter, pUC-UL89P, or the plasmid with a nine amino acid deletion at the N-terminus, pUC-UL89P∆9aa, could complement the replication-deficient genome HCMV-GFPBAC-∆UL89, while the mutant plasmids lacking either the region encoding aa 580-590 or the region encoding aa 590600 were unable to complement the recombinant genome. This indicates that the interaction with the terminase subunit pUL56 is a prerequisite for the function of the HCMV terminase subunit pUL89 during viral DNA replication. Taken together with our earlier observations (19-20) we could clearly demonstrate that both terminase subunits interact with each other and that the interaction domain is localized between aa 580 and aa 600. By complementing the HCMV UL89 null genome, we demonstrated that these amino acids were required for the function in cleavage and packaging of newly synthesized DNA. ACKNOWLEDGMENT We thank T. Stamminger (Institute of Clinical and Molecular Virology, Erlangen) for generously providing antiIE1 antibody (mAb63-27) and IE1 expression plasmid pHM123. E.B. thanks B. Fleckenstein for kind support. REFERENCES 1. Penfold, M. E., and Mocarsci, E. S. (1997) Formation of cytomegalovirus DNA replication compartments defined by localization of viral proteins and DNA synthesis,Virology 239, 46-61. 2. Black, L. W. (1988) DNA packaging in dsDNA bacteriophages, in The Bacteriophages (Calendar, R., Ed), Plenum, New York. 3. Bhattacharyya, S. P., and Rao, V. B. (1993) A novel terminase activity associated with the DNA-packaging protein gp17 of bacteriophage T4, Virology 196, 34-44. 4. Feiss, M., and Becker, A. (1983) DNA packaging and cutting, in Lambda II (Hendrix, R. W., Roberts, J. W., Stahl, F. W., and Weisberg, R. A., Eds.) pp 305-330, Cold Spring Harbor, New York. 5. Becker, A., and Murialdo, H. (1990) Bacteriophage lambda DNA: the beginning of the end, J. Bacteriol. 172, 2819-2824. 6. Fujisawa, H., and Hearing, P. (1994) Structure, function and specificty of the DNA packaging signals in double-stranded DNA viruses, Semin. Virol. 5, 5-13. 7. Feiss, M. (1986) Structure, function and specificity of the DNA packaging signals in double-stranded DNA viruses, Trends Genet. 2, 100-104. 8. Guo, P., Peterson, C., and Anderson, D. (1987) Prohead and DNAgp3-dependent ATPase activity of the DNA packaging protein gp16 of bacteriophage phi 29, J. Mol. Biol. 197, 229-236. 9. Catalano, C. E., Cue, D., and. Feiss, M. (1995) Virus DNA packaging: the strategy used by phage lambda, Mol. Microbiol. 16, 1075-1086. 10. Hang, J. Q., Tack B. F., and. Feiss, M. (2000) ATPase center of bacteriophage lambda terminase involved in post-cleavage stages of DNA packaging: identification of ATP-interactive amino acids, J. Mol. Biol. 302, 777-795. 11. Bhattacharyya, S. P., and Rao, V. B. (1994) Structural analysis of DNA cleaved in ViVo by bacteriophage T4 terminase, Gene 146, 67-72.

Interaction Domain HCMV pUL89 12. Morita, M., Taska, M., and Fujisawa, H. (1993) DNA packaging ATPase of bacteriophage T3, Virology 193, 748-752. 13. Rao, V. B., and Black, L. W. (1988) Cloning, overexpression and purification of the terminase proteins gp16 and gp17 of bacteriophage T4. Construction of a defined in-vitro DNA packaging system using purified terminase proteins, J. Mol. Biol. 200, 475488. 14. Underwood, M. R., Harvey, R. J., Stanat, S. C., Hemphill, M. L., Miller, T., Drach, J. C., Townsend, L. B., and Biron, K. K. (1998) Inhibition of human cytomegalovirus DNA maturation by a benzimidazole ribonucleoside is mediated through the UL89 gene product, J. Virol. 72, 717-725. 15. Krosky, P. M., Underwood, M. R., Turk, S. R., Feng, K. W-H, Jain, R. K., Ptak, R. G., Westerman, A. C., Biron, K. K., Townsend, L. B., and Drach, J. C. (1998) Resistance of human cytomegalovirus to benzimidazole ribonucleosides maps to two open reading frames: UL89 and UL56, J. Virol. 72, 4721-4728. 16. Bogner, E., Reschke, M. Reis, B. Mockenhaupt, T., and Radsak, K. (1993) Identification of the gene product encoded by ORF UL56 of the human cytomegalovirus genome, Virology 196, 290293. 17. Bogner, E., Radsak, K., and Stinski, M. F. (1998) The gene product of human cytomegalovirus open reading frame UL56 binds the pac motif and has specific nuclease activity, J. Virol. 72, 22592264. 18. Bogner, E. (2002) Human cytomegalovirus terminase as a target for antiviral chemotherapy, ReV. Med. Virol. 12, 115-127. 19. Hwang, J.-S., and Bogner, E. (2002) ATPase activity of the terminase subunit pUl56 of human cytomegalovirus, J. Biol. Chem. 277, 6943-6948. 20. Scheffczik, H., Savva, C. G. W., Holzenburg, A., Kolesnikova, L., and Bogner, E. (2002) The terminase subunits pUL56 and pUL89 of human cytomegalovirus are DNA metabolizing proteins with toroidal structure, Nucleic Acids Res. 30, 1695-1703. 21. Scholz, B., Rechter, S., Drach, J. C., Townsend, L. B., and Bogner, E. (2003) Identification of the ATP-binding site in the terminase subunit pUL56 of human cytomegalovirus, Nucleic Acids Res. 31, 1426-1433. 22. Savva, C. G. W., Holzenburg, A., and Bogner, E. (2004) Insights into the structure of human cytomegalovirus large terminase subunit pUL56, FEBS Lett. 563, 135-140. 23. Adelman, K., Salmon, B., and Baines, J. D. (2001) Herpes simplex virus DNA packaging sequences adopt novel structures that are specifically recognized by a component of the cleavage and packaging machinery, Proc. Natl. Acad. Sci. U.S.A. 98, 30863091. 24. Bogner, E., Reschke, M., Reis, B., Reis, E., Britt, W., and Radsak, K. (1992) Recognition of compartmentalized intracellular analogs of glycoprotein H of human cytomegalovirus, Arch. Virol. 126, 67-80. 25. Datsenko, K. A., and Wanner, B. L. (2000) One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products, Proc. Natl. Acad. Sci. U.S.A. 97, 6640-6645. 26. Borst, E. V., and Messerle, M. (2005) Analysis of human cytomegalovirus oriLyt sequence requirements in the context of the viral genome, J. Virol. 79, 3615-3626. 27. Borst, E. M., Hahn, G., Koszinowski, U. H., and Messerle, M. (1999) Cloning of the human cytomegalovirus (HCMV) genome

Biochemistry, Vol. 45, No. 29, 2006 8863 as an infectious bacterial artificial chromosome in Escherichia coli: a new approach for construction of HCMV mutants, J. Virol. 73, 8320-8329. 28. Adler, H., Messerle, M., Wagner, M., and Koszinowski, U. H. (2000) Cloning and mutagenesis of the murine gammaherpesvirus 68 genome as an infectious bacterial artificial chromosome, J. Virol. 74, 6964-6974. 29. Borst, E. M., Mathys, S., Wagner, M., Muranyi, W., and Messerle, M. (2001) Genetic evidence of an essential role for cytomegalovirus small capsid protein in viral growth, J. Virol. 75, 14501458. 30. Smuda, C., Bogner, E., and Radsak, K. (1997) The human cytomegalovirus glycoprotein B gene (ORF UL55) is expressed early in the infectious cycle, J. Gen. Virol. 78, 1981-1992. 31. Andreoni, M., Faircloth, M. Vulger, L., and Britt, W. J. (1989) A rapid microneutralization assay for the measurement of neutralizing antibody reactive with human cytomegalovirus, J. Virol. Methods 23, 157-167. 32. Fleckenstein, B., Mu¨ller, I., and Collins, I. (1982) Cloning of the complete human cytomegalovirus genome in cosmids, Gene 18, 39-46. 33. Salmon, B., and Baines, J. D. (1998) Herpes simplex virus DNA cleavage and packaging: association of multiple forms of U(L)15-encoded proteins with B capsids requires at least the U(L)6, U(L)17, and U(L)28 genes, J. Virol. 72, 3045-3050. 34. Yu, D., and Weller, S. K. (1998) Herpes simplex virus type 1 cleavage and packaging proteins UL15 and UL28 are associated with B but not C capsids during packaging, J. Virol. 72, 74287439. 35. White, C. A., Stow, N. D., Patel, A. H., Hughes, M., and Preston, V. G. (2003) Herpes simplex virus type 1 portal protein UL6 interacts with the putative terminase subunits UL15 and UL28, J. Virol. 77, 6351-6358. 36. Catalano, C. E. (2000) The terminase enzyme from bacteriophage lambda: a DNA-packaging machine, Cell. Mol. Life Sci. 57, 1075-1086. 37. Lin, H., Simon, M. N., and Black, L. W. (1997). Purification and characterization of the small subunit of phage T4 terminase, gp16, required for DNA packaging, J Biol. Chem. 272, 3495-3501. 38. Giesen, K., Radsak, K., and Bogner, E. (2000a) Targeting of the gene product encoded by ORF UL56 of human cytomegalovirus into viral replication centres, FEBS Lett. 471, 215-218. 39. Abbotts, A. P., Preston, V. G., Hughes, M., Patel, A. H., and Stow, N. D. (2000) Interaction of the herpes simplex virus type 1 packaging protein UL15 with full-length and deleted forms of the UL28 protein, J. Gen. Virol. 81, 2999-3009. 40. Koslowski, K. M., Shaker, P. R., Casey, J. T., II, Wilson, T., Yamanaka, G., Sheaffer, A. K., Tenney, D. J., and Pederson, N. E. (1999) Physical and functional interaction between the herpes simplex virus UL15 and UL28 DNA cleavage and packaging proteins, J. Virol. 73, 1704-1707. 41. Giesen, K., Radsak, K., and Bogner, E. (2000b) The potential terminase subunit pUL56 of HCMV is translocated into the nucleus by its own NLS and interacts with importin R, J. Gen. Virol. 81, 2231-2244. BI0600796